THE P53-POSITIVE PHENOTYPE OF BREAST CANCERS - CORRELATION WITH OTHERPARAMETERS AND CONSEQUENCES ON ITS BIOLOGICAL SIGNIFICANCE

One hundred and eighty-one breast cancer specimens were analyzed for n uclear p53 staining by immunochemical methods. There were 123 fine-nee dle cytological specimens and 58 frozen tissue sections of surgical bi opsies. The microscopic evaluation of the staining fitted with a 4 gro up classification. Ninety-one samples (50.6%) were devoid of any stain ing (-), while 42 (23.3%) showed only few stained nuclei (+/-), typica lly around 1%. Thirty-two (17.8%) samples presented with strong nuclea r staining (++) which in practically all cases concerned more than 50% of the nuclei, but a few cases showed staining heterogeneity. A furth er 17 cases (9.4%) presented with nuclear staining which concerned 10- 20% of the cancer cells (+). This four class system was used to compar e p53 expression with other prognostic parameters. A strong inverse co rrelation was observed with steroid hormone receptor content and p53 p ositivity was highly significantly associated with higher S-phase. All but one of the highly positive cases were aneuploid. Twenty-five perc ent (29/120) of the aneuploid tumors were strongly stained and a furth er 10% were considered positive (+). On the other hand, only 5 out of 59 DNA-diploid tumors were considered as + and one ++. The DNA index d istribution according to p53 positivity showed peaks of positivity for hypodiploid, triploid and hypertetraploid values. Negative tumors wer e in all regards similar to those with only few stained nuclei, in par ticular mean S-phases of 2.8 and 3.3% respectively. Altogether, the ty pical strong p53 phenotype concerned a DNA-aneuploid tumor with above median S-phase fraction (mean of 7.1%), negative steroid hormone recep tors and cytoprognostic index III. The p53 positive cases (+), were fr equently steroid hormone receptor positive and had on the average inte rmediate S-phase fractions (4.3%). The proportion of immunochemical po sitivity (27% in our series), is compatible with the published frequen cy of p53 mutations detected in breast cancers, but the differences in the phenotype according to the level of positivity should be further investigated.